In Germany, cycling has experienced a considerably increase in popularity in the past few years. The number of e-bikes (mainly pedelec25 and pedelec45) on German roads is growing steadily as well. As a consequence, investigations of safety relevant behaviour of bicyclists and e-bike riders, such as helmet use or violations (e.g. red light running) are of vital importance. To address such issues, data of the previous Pedelec Naturalistic Cycling Study (Pedelec-NCS), in which (e-)cyclists’ behaviour was observed under natural conditions, was reanalysed. Focus of the Pedelec-NCS was the comparison of conventional bicycles, pedelec25 and pedelec45. The e-bikes and bicycles of 90 participants (49 pedelec25, 10 pedelec45, 31 conventional bicycles) were instrumented with two cameras (forward view and face of the rider) and additional sensors. A total of more than 4,300 trips, with nearly 17,000 km cycled distance was recorded. In the reanalysis, the following four issues were addressed: 1. Relation between helmet use, trip length and cycling speed: One major argument against a compulsory use of cycling helmets are potential safety issues. It has been suggested that cyclists that wear a helmet might compensate for this perceived increase in safety by cycling more risky. This could mean that they cycle faster, and hence increase their risk of a crash. However, evidence for such an assumption, especially under natural conditions, has been lacking so far. The role of additional factors that might have an influence, such as the length of a trip, was not fully clear as well. Therefore, the data of 85 Pedelec-NCS participants was used to investigate the relationship between helmet use, trip length and cycling speed. Using the face view videos, more than 3,700 trips were reviewed to annotate whether the riders wore a helmet or not. This annotation was then connected to information on trip length and cycling speed. Overall, a helmet usage quote of 58% was found, i.e. the participants wore a helmet in more than half of the collected trips. Pedelec45 riders showed the highest usage rate, riders of conventional bicycles the lowest. The analysis of factors influencing cycling speed showed that trip length had a much stronger effect on the riders speed than the question of whether they had worn a helmet or not, which indicates that helmet use only has a minor role (if any at all) for the speed of bicyclists and e-bike riders. The suggestion of a relevant behavioural adaptation as a result of the potentially increased safety because of the use of a helmet does not appear to be supported. 2. Traffic light violations (red light running): Although violations of traffic lights are among the most common offenses committed by cyclists, there is hardly any information on the specific circumstances of these violations (e.g., infrastructural characteristics, specific behavioural patterns). Likewise, information on potential differences between riders of conventional bicycles and e-bikes (pedelec25 and pedelec45) with regard to their propensity to violate traffic lights is lacking. The data of 88 Pedelec-NCS participants was used to address these questions. In total, nearly 8,000 red light encounters were identified, with a red light running rate of 17%. In addition, in 5% of all red light encounters, participants changed the used infrastructure to avoid the red light, so in total, a traffic light violation was observed in more than 20% of all cases. When turning right, violations were even more common than rule abiding behaviour. There were no differences between the riders of conventional bicycles and the riders of the different e-bikes. In nearly three quarters of all cases, riders appeared to run the red light without any visible braking or stopping prior to the violation. The highest violation rates were observed at T-junctions when turning out, as well as at roads sections without intersection (e.g., traffic lights only for pedestrians to cross). It can be suspected that situations in which the overall traffic situation can be easily assessed by the cyclist, and in which crossing traffic is rare, might somewhat encourage the cyclist to violate the traffic light. 3. Illegal use of sidewalk instead of carriageway: The illegal use of sidewalks often results in conflicts and even collisions with pedestrians. Assessing potentially motivating factors on a situational level appears to be a logical step towards eventual countermeasures. The data of 81 Pedelec-NCS participants, all of which were observed to at least once ride on the sidewalk instead of the roadway, was used to for this analysis. In the subsequent video annotation, potentially motivating factors for the illegal use were identified for each case. Factors that could be summarised as focussing on efficiency were most prominent. Especially the potential motive of maintaining speed was observed repeatedly, which is not surprising, given that maintaining speed also means avoiding the need to accelerate the bike again. Factors that focused on safety, such as using the sidewalk to avoid interacting with a particular motorised road user, or to avoid a busy street in general, were observed as well. It can be assumed that infrastructural characteristics played at least a partial role in the occurrence of this type of behaviour. 4. (Illegal) Riding against traffic on the roadway or cycling infrastructure: Similar to the illegal use of the sidewalk, potentially motiving factors for riding against traffic were assessed. The data of 46 Pedelec-NCS participants, all of whom were observed to at least once ride against traffic on the roadway or cycling infrastructure, was used to for this analysis. Again, potentially motivating factors for this behaviour were identified through video annotation. It appeared that unfavourable infrastructural characteristics, such as a lack of cycling infrastructure in the direction of travel, difficulties in accessing the “correct” infrastructure (e.g., because of road barriers) or the road/infrastructure surface (potholes etc.), were highly prevalent in the observed cases. Improvements of this infrastructure appear to be logical step in order to stop this illegal behaviour and, as a result, reduce crash risk. (Author/publisher)
Samenvatting